The Indian Robo-Rush

FIFTEEN YEARS AGO, when still an engineering student, Gokul NA co-authored a paper on designing a robot that could climb areca nut trees. It was a college project but he still had a larger ambition—to come up with a mar­ketable product. It soon ran into hurdles. For one, the field that he was interested in—robotic arms—did not have demand from industry and that was despite them being around from the 1950s. “It is still only a $50 billion market, of which only $16 billion is sale of the robotic arms, the rest is services around it. Why is this so small a market was the question. Which meant, somewhere, the utility for the robotic arms was very minimal. They were not sufficiently replacing the human ability to do or manipulate tasks,” he says.

Human hands adjust to changes in the environment by instinct. If an object that we picked up earlier changes position—say a bottle that is now upside down—there is nothing extra required to flip it back. “We are doing it subconsciously. These are traits not taught to us,” he says. Robots haven’t been that way. Industrial automation, such as factory production lines, largely uses them, but these were limited to spe­cific tasks performed in a particular way. Gokul’s objective became to come up with robots that could visualise like humans and be flexible in the tasks they can do. He did a corporate stint and then, along with a part­ner, launched a startup. It was a tough field to venture into because robots operating on vision had a poor performance record.

“Customers would outright tell you that if you’re trying to solve my automa­tion problem using vision, please don’t. Just use some mechanised mechanism to solve it. We don’t want to be stuck with vision not working half the time,” he said. Gokul realised the problem was with the way vision was conceived in technology. It was mainly for identification, photog­raphy, etc. Vision, as a means to object manipulation, had never been in the equation. It would therefore need a fresh approach. That is what he set out to do.

His company CynLr has now created a semi-humanoid robotic arm, called Cyro, which functions at its core by mimicking portions of the brain. It looks futuristic, with three metal appendages that ema­nate from a trolley. It has taken him a little under 20 years from when he first began to think of robots as a student, but is now at rollout point. Multinationals like the US automobile company General Motors, and the Japanese automotive components manufacturer Denso have partnered with him. “We have released the first version of the product for customer trials. We have customer engagements where we are do­ing trials in-house. We are at a pilot stage with General Motors and Denso. We are also onboarding a few other manufactur­ing customers this year,” he says.

What is the advantage of such a robot? For one, it can radically improve manufac­turing. Take a car company, for example, that wants to set up a production line in India. Once it is automated, changing the process is difficult because it requires rede­signing the hardware. Say, it spends $300 million to set up and needs to produce and sell 30,000 cars a month to be profitable. “But if the market demands four different models instead of one, producing all four on the same line becomes a challenge. Our solution centres on the concept of a universal factory, which we propose to customers to rethink how factories are designed,” he says. So like human beings, the same robot can change workstations and do different tasks. For the automotive parts company Denso, they showed how it worked. It involved a high-mix, low-vol­ume production line where different parts are assembled in each shift. “Our robots can handle gears in one shift, rotors in the next, and wiring in another, all without changing the hardware. This means the platform is fixed, and instead of building unique machines for each station, the entire line can replicate the same adapt­able system, enabling flexibility and scal­ability in manufacturing,” he says.

It has been evident for a long time that robotics was the future of manufactur­ing, but it is only now that there has been a sudden explosion of interest. It is as if a critical mass is inevitably approaching. India, for instance, is now a major market for industrial robots. Last year, when the World Robotics 2024 report, presented by the International Federation of Robotics (IFR), was released, it noted that India is sev­enth in annual robot installations. Sales for 2023 touched a new record of 8,510 units, an increase of 59 per cent. Their press release said: “Installations in the automotive in­dustry jump 139% to 3,551 units in 2023. Both car manufacturers and parts suppliers contributed to this development. This seg­ment accounts for a market share of 42%. The general industry in India is led by the rubber and plastics industry with 408 units, up 17%. Installations in the metal industry remained steady at 329 units in 2023. The long-term potential of robotics in India be­comes clearer when compared to China, which has a similar population size: An­nual installations in China reached 276,288 units in 2023. This is 32 times higher than the current peak level in India.”

In 2022, Reliance Industries, India’s biggest company, bought a majority stake in the robotics startup Addverb for $132 million. Addverb’s founders had worked in large enterprises like Asian Paints be­fore they saw the opportunity for an au­tomation provider in India. Prateek Jain, a co-founder and chief operations officer, says that the early years after their 2016 launch were difficult because businesses had to be convinced to embrace robotics. “A defining moment was when we suc­cessfully deployed our first largescale au­tomation project for HUL [Hindustan Uni­lever], proving that homegrown robotics solutions could match global standards. That win cemented our belief that India could be a global leader in robotics,” he says over email.

India, according to him, is now at an inflection point in robotics adoption. The sectors he lists where it is becoming increasingly used range from manufac­turing and automotive, to healthcare, agriculture, and logistics. “In health­care, our medical cobot [companion robot] like Heal is streamlining hospital operations with an innovative approach to rehabilitation and imaging, while exoskeletons are aiding in rehabilita­tion and reducing caregiver fatigue. In agriculture, robotic weeders, drones, and fruit-picking systems though not wide­spread are gaining traction through research and startups,” he says.

Addverb is now taking a next leap with plans for humanoids. These are robots that look and interact like humans, some­thing that is pegged to move from science fiction into the real world. Tesla, the electric vehicle behemoth of Elon Musk, for instance, has a humanoid called Optimus in the works. Nvidia, whose GPU chips are heralding the artificial intelligence (AI) revo­lution, announced a foundation model, or a software framework, for such robots at their developer conference in March. Its founder and CEO Jensen Huang said the time when humanoid robots could be seen mixing in society was imminent. Reuters reported, “Speaking to a group of journal­ists after the speech, Huang was asked what signs would show that AI had become ubiquitous. Among other answers, Huang said it may be ‘when, literally, humanoid robots are wandering around, which is not five years away. This is not a five-years-away problem, this is a few-years-away problem.’ The manufacturing industry would likely adopt humanoid robots first because that industry has well-defined tasks that robots can handle in a controlled environment, he said. ”

According to Jain, humanoid robot­ics in India can have use-cases across the spectrum. In healthcare, they could provide support for patients and the elderly; retail and hospitality industries could use them for customer service and visitor assistance; in assembly lines for manufacturing; in defence for security, surveillance, and administrative roles. Addverb is still in the starting stages of research and development with humanoids. “We’re actively evaluating the feasibility and use cases of humanoid robots, with initial timelines likely taking shape by the last quarter of 2025,” Jain says.

THE ROBOTICS ECOSYSTEM in India is just beginning, limited to a handful of startups, but some are doing interesting work. GenRobotics Innovations, a Kerala startup, for example, has a scavenger that cleans sewer manholes. Bengaluru-based Ati Motors, which has autonomous moving robots, raised $20 million in investment early this year for its expansion plans. An unusual robot is from the stables of the Indian Space Research Organisation (ISRO) and related to Gaganyaan, ISRO’s human spaceflight programme slated for 2026. But before that, on a testing flight, it will send a humanoid robot called Vyommitra into space. A paperpublishedthismonthbytheObserver Research Foundation titled ‘AI in Space Operations: Opportunities and Challeng­es’ said, “This robot will test the Geosyn­chronous Satellite Launch Vehicle Mark III [GSLV Mk III] rocket and ensure its safety for human travel, tracking changes in the crew module during spaceflight and return.”

The potential market for Indian startups can be seen in their future plans. Addverb says its goal is to grow at a rate of 60-80 per cent annually over the next five years and become a billion-dollar revenue company, with 70 per cent of its revenues coming from overseas markets. “We’ve already made strong headway in US, Europe and Asia, and we are planning to expand in Africa as well,” says Jain.

CynLr currently has around 60 employ­ees and Gokul says that in five years, they aim to break even by targeting $50-100 million in revenue. “It’s hard to predict exactly where we’ll land, but that’s the goal we’re working towards. We’re also expanding globally, both in terms of talent and customers. Hopefully, we’ll be selling close to 1,000 robots annually, which would help us reach that revenue target. To support selling 1,000 robots an­nually, we’d need a team of about 1,000 to 1,300 members,” he says.

There are challenges. For one, finding good technological talent in India is diffi­cult because many go to work abroad. But Gokul thinks India has an opportunity to be at the forefront of this moment. He says, “The robotics industry is still nascent, and if we step in now, we could own the space, much like we’ve done in software. We have the technological skills and human resources in India to make this happen. We need to train and groom talent, build infrastructure, and create a supply chain to capture the robotics market early.”

Addverbartificial intelligenceCYROGokul NAhumanoid helpersPrateek Jainroboticstechnology